The invention relates to an electromagnetic flowmeter arrangement, having a measuring tube, a coil arrangement for generating a magnetic field substantially perpendicular to the direction of flow through the measuring tube, an electrode arrangement substantially perpendicular to the direction of flow and to the magnetic field, an evaluating device and a testing device.
An arrangement of that kind is known from GB 2 309 308 A. Here, to examine or test the measuring tube with its evaluating device, the normal connection between the measuring tube and its evaluating device is interrupted. Then an external measuring circuit is connected to the evaluating device and to the measuring tube. During the test, the flow through the measuring tube is not measured. The last value to be measured is retained by the evaluating device. The measuring circuit first of all determines the ohmic resistance of the coil arrangement by applying a voltage to the coil arrangement. As soon as the ohmic resistance has been determined by ratio forming, the voltage across the coil is set to zero and the exponentially decaying coil current is monitored, which then enables the inductance of the coil to be determined. The comparison between desired and actual values of ohmic resistance and inductance gives information about possible changes to the flowmeter arrangement which necessitate a re-calibration. Calibration is then carried out by using new calculation values.
The expense involved in a test is relatively high. As a consequence, there is a risk that a test will be carried out only at relatively large intervals. There is also the danger that the interruption and reconnection of leads will introduce errors, which will have an adverse effect on the actual measurement, for example, through inaccurate or false measurement values. This can become problematical in particular when the flowmeter is used to account for consumption.
The problem underlying the invention is to be able to carry out testing of the flowmeter arrangement in a simple manner.
The problem is solved in an electromagnetic flowmeter arrangement of the kind mentioned in the introduction in that the testing device comprises first means for applying voltage to the coil arrangement without generating a magnetic field, and second means for determining an ohmic and/or inductive and/or capacitive coupling between the coil arrangement and the evaluating device.
During examination (in the following also described as xe2x80x9ctestingxe2x80x9d), electrical relationships are therefore created, which come relatively close to those in operation. To be precise, voltage is applied to the coil and signals can be taken from the electrode arrangement. Unlike normal measuring operation, these signals cannot have been generated by the magnetic field and the fluid flowing through, because precautions have been taken to ensure that no magnetic field is generated. If signals do occur, these can only be attributable to an electrical coupling between the coil arrangement and the evaluating device. If this coupling changes, then this is an indication that the flowmeter arrangement as a whole has changed, so that possibly a calibration is needed. Should this coupling not have changed, one can assume that a calibration performed originally continues to be valid. The coupling can be expressed by different physical variables. The ohmic, inductive and capacitive coupling or just one or two kinds of this coupling can be monitored. The time needed to determine this coupling is not generally very long. Testing can therefore also be carried out during normal measuring operation, which need only be interrupted very briefly for that purpose.
The second means are preferably formed by the evaluating device or are integrated therein. By means of the evaluating device the signals coming from the electrode arrangement can therefore be detected and processed in the same way as the signals determined during measuring. The evaluating device merely has to be informed that the signals present here are test signals and not measuring signals. Further processing can depend on the test signals.
The testing device advantageously comprises a timer, which initiates testing at the latest after expiry of predetermined time intervals. One is not then dependent on carrying out testing from time to time oneself or by maintenance personnel. On the contrary, testing is effected automatically at the latest at the end of predetermined time intervals. These test intervals can, of course, also be shortened.
The voltage is preferably in the form of a controlled alternating voltage. Accordingly, the coil arrangement has different voltage potentials applied to it and for the different voltage potentials the corresponding couplings can be examined. The danger that an error will remain undiscovered because it occurs only in a specific operating state of the voltage is relatively small, because all voltage levels occurring in operation are passed through.
In this connection, it is especially preferred that the voltage is formed by a supplementary voltage. Such a xe2x80x9cboostxe2x80x9d voltage is an increased voltage that is already present in many flowmeters to accelerate the build-up of the magnetic field. This higher voltage renders a coupling that may already be present more easily recognizable.
It is especially preferred in this connection that the voltage during testing has the same parameters as during measuring. In other words, the xe2x80x9cboostxe2x80x9d voltage that is present anyway can be used to carry out testing as well. During measuring and during testing the xe2x80x9cboostxe2x80x9d voltage then has the same amplitude and the same frequency.
The voltage is preferably generated by an H-bridge, which has in each branch a controlled switch, the H-bridge being active crosswise during measuring and sidewise during testing. During measuring, the diagonal branches of the H, in the middle of which the coil arrangement is connected, are operated in a manner known per se so that a current can flow in one or other direction through the coil arrangement. An alternating field is generated in this way in the coil arrangement. During testing, the same H-bridge can be used, the difference being merely that the bridge is no longer operated crosswise, but in each case a half of the bridge branches, namely the branches that are connected to the same voltage potential, are closed. It is therefore possible in a simply way to apply voltage to the coil arrangement without generating a magnetic field. Current flow is prevented.
The coils of the coil arrangement are preferably short-circuited during testing. Conditions in the coils of the coil arrangement are therefore the same and testing is simplified.
The short-circuited coils are advantageously connected alternately to a voltage source and to earth during testing. Leakage currents from the coil arrangement to the evaluating device and vice versa can therefore be detected.
It is also advantageous for the testing device to have a memory device in which values determined during testing and/or desired values are saved. The desired values can originate, for example, from the first or from the last valid calibration. The values ascertained during testing can then be compared with the values in memory. Variations can be used to assess whether the flowmeter arrangement is still operating sufficiently reliably or not. The memory device can also be used, however, to save one after the other a certain number of test values and as it were to record the xe2x80x9chistoryxe2x80x9d of the flowmeter arrangement. Of course, past test values can also be xe2x80x9ccompressedxe2x80x9d and, for example, their mean value and their range of variation can be saved in memory, these variables being updated at each test and then saved again.
The testing device preferably determines a reliability coefficient from the spread of values determined during testing and repeats the tests in dependence on the reliability coefficient. In this connection, the assumption is that testing should take place more frequently when it appears that the individual values are widely spread. If, on the other hand, no great variations are discernible, then there can be a longer period of waiting to the next test. Here, an upward limit is set by means of the predetermined time intervals, which have been mentioned above.